Multi-element decision-making method and device based on scheme attributes and expert attributes

The invention provides a multivariate decision-making method and device based on scheme attributes and expert attributes, belongs to the technical field of scheme decision-making, and designs a brand-new multivariate decision-making model which can effectively learn and integrate experience knowledge of different experts, gradually realize quantitative and qualitative combination of evaluation attributes and improve decision-making efficiency. It is ensured that finally obtained expert group decision evaluation is more scientific and comprehensive; aiming at the defect of weight configuration in a traditional method, a neural network technology is innovatively introduced, and expert weights and attribute weights are automatically learned and optimized by training a neural network. The problems that pure subjective weighting is unstable and an objective weighting method neglects the view angle of experts are solved, dynamic adjustment and optimization of weight configuration are achieved, and the accuracy and